Internalized Stigma, Insight, Social Anxiety and Quality of Life in First-Episode Psychosis: A Cross-Sectional Study

1. Introduction

Stigma is a characteristic or attribute that a person possesses and is associated with negative qualities in a specific social context. It is devalued, leading to negative attitudes towards the stigmatized person (Crocker et al. 1998). 87% of persons with severe mental illness have reported stigma experiences (Wood et al., 2017), and people with schizophrenia belong to one of the most stigmatized social groups (Brohan et al., 2010). They are often considered dangerous, unpredictable, or unlikely to recover (Wood et al., 2014). Patients consider stigma as more disabling than the disease itself, as a “second illness” (Morrison et al., 2016).

Thornicroft (2007) has focused on three psychological aspects of stigma: knowledge (misinformation, ignorance), attitude (prejudice) and behavior (discrimination against the stigmatized).

Stigma can be distinguished in: 1. Perceived stigma (the conviction that society has a negative opinion of a specific characteristic a person has), 2. Experienced stigma (verbal or physical abuse, bullying, discrimination in education and employment opportunities, parenting or interpersonal relationships because of the stigmatized attribute, and 3. Internalized stigma (the belief that the negative stereotypes apply to themselves) (Yanos et al., 2008). Almost half of the patients with psychosis report moderate to high levels of internalized stigma, and two-thirds report discrimination experiences due to mental illness (Brohan et al., 2010).

Self-stigma is associated with reduced self-esteem, depression, and social anxiety (Birchwood et al., 2007), fewer educational and employment opportunities, fewer interpersonal relationships, and social isolation (Morrison et al., 2016; Brain et al., 2014; King et al., 2007). It can also affect the subjective community belonging (Berry et al., 2018).

Stigma increases non-adherence to antipsychotic medication, bringing about increased relapse rates, frequent hospitalizations, and poor prognosis (Haddad et al., 2014). Patients with psychosis reported that in order to avoid the stigma of taking antipsychotics, they hide their medication from others and take it when they are alone (Jerkins & Carpenter-Song, 2009).

Furthermore, stigma is correlated with suicidality in people with psychosis. More than half of persons with psychotic disorders who attempted suicide report that stigma played a role in their decision (Eagles et al., 2003).

Studies conducted on hospitalized patients with psychosis have demonstrated that stigma was related to all aspects of the hospitalization including admission, staff behavior and the post-discharge period in the community (Wood et al., 2016). Hospitalization-related stigma may be a risk factor for suicide attempts in the immediate post-discharge period (Schomerus et al., 2015).

Although many studies have been conducted in patients with chronic schizophrenia, only a few deal with stigma in First Episode Psychosis (Ho et al., 2018).

Persons who have a first psychotic experience and a first contact with psychiatric health services may be more vulnerable to stigmatizing attitudes and to stereotypes. However, anti-stigma interventions may be more effective when applied early in this population (Ho et al., 2018).

The aim of this study was to examine the relationships between internalized stigma, insight, social anxiety, and quality of life in first-episode psychosis.

Despite increasing research on stigma in chronic schizophrenia, evidence focusing specifically on first-episode psychosis remains limited, notably about the combined role of insight, social anxiety, and quality of life within early recovery stages. Understanding these relationships may help identify modifiable targets for early intervention services.

2. Methods

This was an observational cross-sectional study conducted within a specialized Early Intervention in Psychosis service, assessing psychosocial outcomes six months following stabilization after a first psychotic episode.

2.1. Participants

Patients were recruited from the Early Intervention in Psychosis Unit, Department of Psychiatry, University Hospital of Ioannina, Greece, from October 2019 till June 2023. Ninety patients were enrolled in the study according to the following inclusion and exclusion criteria.

Inclusion criteria were: a. experience and recovery from a first episode of psychosis, b. diagnosis of brief psychotic episode, schizophrenic form disorder or schizophrenia following the criteria of the Diagnostic and Statistical Manual of Mental Disorders-V (DSM-V) (American Psychiatric Association, 2013), c. members of the outpatient early intervention service for at least six months, and agreement in a written informed consent.

Exclusion criteria were: a. history of alcohol and/or substance abuse, b. refusal to give written informed consent, and c. inability to understand and write the Greek language for any reason.

2.2. Materials and Methods

2.2.1. Internalized Stigma for Mental Illness Scale

The Internalized Stigma for Mental Illness Scale (ISMI) is a self-report questionnaire comprising 29 items rated from 1 (strongly disagree) to 4 (strongly agree).ISMI comprises 5 subscales assessing alienation (items 1,5,8,16,17,21), stereotype endorsement (items 2, 6, 10, 18, 19, 23, 29), discrimination experience (items 3,15, 22, 25, 28), social withdrawal (items4, 9,11,12,13, 20) and stigma resistance-reverse stigma (items 7,14,24, 26,27).

It possesses high internal consistency in persons with severe mental disorders, including psychosis. A mean score for the first four subscales was used, as the fifth subscale assesses not stigma but a reaction to it. The questionnaire lacks a cut-off point, with higher scores indicating greater stigma (Paraskevoulakou et al., 2017; Ritsher et al., 2003). We further categorized participants with high indications for InternalizedStigma based on a score greater than 50 points on the total ISMI scale.

2.2.2. World Health Organization Quality of Life Assessment-BREF

The World Health Organization Quality of Life Assessment-BREF (WHOQOL-BREF) consists of 26 items that cover the following four domains: physical health, mental health, social relationships and environment. All items are rated using a 5-point Likert scale. According to the WHO guidelines, all scores were transformed to a 1-100 scale (WHOQOL Group, 1999). Higher scores indicate a better quality of life (Ginieri-Coccossis et al., 2012).

2.2.3. Liebowitz Social Anxiety Scale

Liebowitz Social Anxiety Scale (LSAS) comprises 24 items assessing fear/anxiety and avoidance in different social situations. Each item assesses both variables. Participants are asked to rate their fear/anxiety on a four-point scale (from 0=none to 3=severe) (LSAS-anxiety subscale) and avoidance also on a four-point scale (from 0=never to 3=usually) (LSAS-avoidance subscale). The total score represents the sum of the scores obtained on both subscales (Liebowitz, 1987; Caballo et al., 2019). The Greek translation of LSAS possesses a very high internal consistency (Cronbach’s alpha: 0.97).

2.2.4. Schedule for the Assessment of Insight-Expanded Version

Insight was assessed using the Schedule for the Assessment of Insight-Expanded version (SAI-E). SAI-E assesses the recognition of having a mental illness (items 1-6), the ability to re-label symptoms as pathological (items 7-9) and the compliance with treatment (items A and B).

Items 1-6 (A and B) are rated from 0 to 2, and items 7-9 from 1 to 4, with the total score ranging from 0 to 28. Although SAI-E lacks a cut-off point, higher scores indicate better insight (Kemp & David, 1997; Konstantakopoulos et al., 2013).

2.2.5. Positive and Negative Syndrome Scale

The psychopathology of each participant was estimated using the Positive and Negative Syndrome Scale (PANSS) at two time points (admission to the hospital and six months after discharge). We assessed psychopathology at two time points, assuming that the severity of symptoms, especially of positive symptoms, at the onset of psychosis may impact stigma levels assessed after their remission. Stigma, insight, social anxiety, and quality-of-life measures were administered at the six-month follow-up assessment, after clinical stabilization.

We used the Greek translated/validated versions for the above mentioned scales (Paraskevopoulou et al., 2017; Ginieri-Coccossis et al., 2012; Konstantakopoulos et al., 2013; Lykouras et al., 1997).

2.3. Ethical Issues

The Ethical Committee of the University Hospital of Ioannina, Greece, approved the protocol

2.4. Data and Statistical Analysis

To reduce the risk of type I error, analyses were guided by predefined conceptual hypotheses rather than exploratory testing.

Data were collected and analyzed with SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics were computed; categorical variables were reported as frequencies and percentages, while continuous variables were reported as ranges, means, and standard deviations (SD). The Pearson product-moment correlation coefficient assessed the linear association between the different scales and subscales. Mann-Whitney tests were used to investigate possible differences between genders across all scales. Univariable linear (for ISMI and as WHOQOL-BREF outcomes) or logistic (for high indication of Internalized Stigma) regression analyses were performed using the following candidate predictor variables: age at entry, gender, marital status, education, medication, children, employment, main residence, psychopathology initially and six months after discharge (PANSS baseline, PANSS at six months’ follow-up), duration of untreated psychosis (DUP), insight (SAI-E),social anxiety and avoidance (LSAS Anxiety, LSAS Avoidance, LSAS Total). For WHOQOL-BREF, the ISMI score was further used as a predictor variable, assuming that stigma is more likely to affect one’s quality of life. Multivariable regression models were constructed, adjusting for age, gender, and variables showing potential association in univariable analyses (p<0.10), consistent with established epidemiological modelling approaches.

The study was reported in accordance with STROBE guidelines for observational studies.

3. Results

The primary finding was the consistent association between higher internalized stigma and poorer quality of life across all WHOQOL-BREF domains.

The sample consisted of 90 patients (55 males, 35 females) with a mean age of 31.8±8.6 years: (18-52). The sample’s baseline characteristics are presented in Table 1. Most of the patients were single (n=72; 80%) and unemployed (n=64; 71.1%). Most participants were diagnosed with schizophreniform disorder (n=78; 86.7%), 8 (8.9%) with brief psychotic episodes, and 4 (4.4%) with schizophrenia. The medications used were risperidone (n=24; 26.7%), clozapine (n=34; 37.8%), and olanzapine (n=32; 35.6%) (Table 1). The mean±SD duration of untreated psychosis was 13.4±6.2weeks (Table 2).

The total mean±SD ISMI score was 50.9±7.9 (range: 29-84). 51 of 90 (56.6%) patients had a total ISMI score greater than 50 and were considered to indicate Internalized Stigma. No statistically significant differences between males and females were observed for any of the scales SAI-E, ISMI, WHOQOL-BREF, LSAS total scores or their subscales.

The mean±SD PANSS-total (PANSS-t) score at baseline was 76.1±7.2 and was reduced to 44.9±6.3 at 6 months of follow-up (p<0.001) (Table 2). A statistically significant difference between males and females was found for the subscales PANSS-n (p=0.024) and PANSS-t (p=0.028) at baseline, but not for the subscales PANSS-p and PANSS-g at baseline or any of the PANSS subscales after 6 months of discharge.

The patients’ scores in the utilized scales and subscales are presented in Table 3. Specifically, across the 90 patients, the mean±SD for insight, measured by SAI-E, was 20.3±4.5 (range: 3-28). Patients reported moderate to good quality of life as measured by the WHOQOL-BREF domains (Table 3).

With regard to the sample’s anxiety and avoidance scores on the LSAS, the mean±SD for anxiety was 3.3±8.7 (range: 0-44), while the mean±SD for avoidance was 2.9±7.6 (range: 0-44).

Supplementary Table S1 presents the correlation matrix for the SAI-E, ISMI, WHOQOL-BREF, and LSAS, along with their subscales. As expected, the multiple subscales within each scale were not completely independent, and we also observed several statistically significant correlations among scales and subscales. While there was no statistically significant linear correlation between SAI-E and ISMI, their subscales presented overall small to moderate correlations. SAI-E subscale “recognition of having a mental illness” showed positive correlations with the total ISMI scale (r=0.39, p<0.001) and the subscales: “Alienation” (r=0.35, p=0.001), “Discrimination experience” (r=0.30, p=0.005), and “Social withdrawal” (r=0.40, p<0.001) and a slight negative correlation with ISMI subscale “Stereotype endorsement” (r=-0.25, p=0.017).

The ISMI scale showed statistically significant, moderate, linear associations with the LSAS-anxiety (r=0.40, p<0.001), LSAS-avoidance (r=0.42, p<0.001), and LSAS total score (r=0.41, p<0.001) subscales. The ISMI total score showed a statistically significant, negative, moderate-to-strong linear correlation with the WHOQOL BREF scales, ranging from r = -0.47 to r = -0.61 (all p < 0.001).

Furthermore, the first four ISMI subscales were negatively correlated with the WHOQOL-BREF subscales, with correlation coefficients ranging from r = -0.19 (WHOQOL-BREF “Social relationships” and ISMI “Alienation”, p = 0.06) to r = -0.56 (WHOQOL-BREF “Social relationships” and ISMI “Discrimination experience”, p < 0.001). In contrast, we observed positive small to moderate linear associations between ISMI “Stigma resistance” subscale and WHOQOL-BREF subscales ranging from r=0.13 (“Physical health”, p=0,205) to r=0.32 (“Mental health”, p=0.002).

SAI-E total score showed small negative correlations with WHOQOL-BREF”Physical health” (r=-0.28, p=0.008) and “Mental health”(r=-0.23, p=0.027). SAI-E subscale “Recognition of having a mental illness” was moderately and negatively associated with WHOQOL-BREF subscales “Physical health” (r=-0.42, p<0.001) and “Mental health” (r=-0.39, p<0.001) and showed a small negative correlation with “Environment” (r=0.25, p=0.017). SAI-E subscale “Compliance with treatment” was positively associated with WHOQOL-BREF subscale “Mental health” (r=0.21, p=0.049) and “Environment” (r=0.25, p=0.016).

The WHOQOL-BREF subscales “Physical health”, “Mental health” and “Social relationships” showed statistically significant negative linear correlations of small to moderate size with the LSAS scale and subscales. No significant correlation was observed between the LSAS total score and subscales and the WHOQOL-BREF “Environment” subscale.

We did not observe a significant linear association between the SAI-E and ISMI scales (or their subscales) and PANSS-t at baseline or at 6 months of follow-up. Only the ISMI “Alienation” subscale presented a slight negative correlation with the scale PANSS-p at baseline (r=-0.27, p=0.011).

Regarding predictors of Stigma for Mental Illness, in the univariable models, employment status was inversely associated with the ISMI score (p=0.037). LSAS-Anxiety and LSAS-Avoidance were positively associated (both p<0.001), with an indication of a positive association with SAI-E (p=0.064); however, none of these variables remained statistically significant in the multivariable model (Supplementary Table S2). When we categorized the ISMI score to high (>50) or low (≤50) levels if indication of Internalized Stigma, medication, SAI-E, LSAS-Anxiety, and LSAS-Avoidance were identified as possible predictors for the univariable models, but in the multivariable models only medication remained statistically significant (Odds ratio [OR] for Risperidone vs Clozapine = 0.12, 95% confidence interval [CI]: 0.03 to 0.48, p=0.003), while there was an indication of a positive association with SAI-E (OR per unit increase of SAI-E = 1.12, 95% CI: 0.99 to 1.27, p=0.067), indicating that higher levels of insight may lead to higher levels of Internalized Stigma (Supplementary Table S3).

As for predictors of quality of life, marital status, having children, employment status, residence, medication type, DUP, SAI-E, ISMI, and LSAS-Total were univariably associated with at least one domain of the WHOQOL-BREF (p<0.10) (Table 4). In the multivariable model, only ISMI remained statistically significant across all WHOQOL- BREF domains, with a 1 unit increase in the ISMI scale being associated with -0.73 units in WHOQOL-BREF Physical health (95% CI: -1.09, -0.36), -1.03 units in WHOQOL BREF Mental health (95% CI: -1.44, -0.62), -0.78 units in WHOQOL-BREF Social relationships (95% CI: -1.38, -0.18), -0.68 units in WHOQOL-BREF Environment (95% CI: -0.99, -0.38) (Table 5).

4. Discussion

The present study finds internalised stigma as a central psychosocial construct in first-episode psychosis, independently associated with reduced quality of life and increased social anxiety. This evidence supports the view that stigma is not simply a social consequence of psychosis but a clinically relevant component influencing early recovery progressions. Systematic assessment of internalised stigma during early psychosis care may therefore help identify patients at risk for poorer psychosocial recovery despite symptomatic improvement.

Examination of clinical and psychosocial characteristics further contextualises the role of internalised stigma during early recovery. Beyond symptom stabilisation, participants showed relatively preserved insight and quality of life alongside moderate levels of internalised stigma, suggesting that psychosocial burden persists even during early recovery phases. These findings align with contemporary recovery-oriented models, which highlight subjective well-being and stigma reduction as central determinants of functional recovery beyond symptom remission.

Internalized stigma was reported by 56.6% of participants, with 36.6% demonstrating moderate-to-high stigma levels (ISMI >60) and 1.1% very high stigma severity (ISMI >70).Ho et al. (2018) reported that, in a sample of 136 persons with FEP, 36.8% had moderate to high self-stigma levels at 3-year follow-up.Together, these outcomes indicate that internalised stigma is already well established early in the course of psychosis rather than emerging solely as a consequence of chronic illness.

Internalised stigma was strongly associated with poorer quality of life across domains, supporting its function as a central determinant of subjective recovery.Perceived stigma is considered a key factor causing poor quality of life in persons with schizophrenia (Livingston & Boyd, 2010). Wang et al. (2016) report in a sample of 128 Chinese patients with schizophrenia a strong reduction in quality of life in the domains of psychosocial and motivation/energy one month after discharge from the hospital. Park et al. (2013) report that higher internalized stigma is correlated with poorer quality of life in the Satisfaction with Family and Satisfaction with Social Relations subscales, but not in the Family Contact and Social Contact subscales. They conclude that internalized stigma affects satisfaction, yet not the quantity of contact.Higher insight levels were associated with increased odds of elevated internalised stigma, while greater stigma independently predicted poorer quality of life after adjustment for covariates. Collectively, these observations reinforce the role of self-stigma as a central mediator between illness experience and subjective well-being in early psychosis.

Chio et al. (2018) report that good insight at baseline was correlated with lower life satisfaction at 6-month follow-up when self-stigma was high, but with better quality of life at one-year follow-up when self-stigma was low. These outcomes indicate that recovery processes may be adversely affected when increased illness awareness is not accompanied by reductions in self-stigma.Buchman-Wildbaum et al. (2020) report in a sample of 200 patients with mental illness (53 of them having a diagnosis of schizophrenia) that better insight was correlated with higher internalized stigma, lower self-esteem and increased experience of shame. These data support the hypothesis of an “insight paradox” (Lysaker et al., 2007), namely that having a good insight into one’s illness may also have negative or even detrimental consequences. Our findings support the “insight paradox” framework, whereby increased illness awareness may coexist with increased self-stigmatization, possibly compromising subjective recovery despite clinical stabilization. Previous research has linked higher levels of insight with depression, anxiety, poorer quality of life, and reduced self-esteem (Buchman-Wildbaum et al., 2020). In this study, the SAI-E subscale “recognition of having a mental illness” was positively correlated with the total ISMI scale and the subscales “Alienation”, “Discrimination experience” and “Social withdrawal” while the SAI-E subscale “ability to re-label symptoms as pathological” was negatively associated with the ISMI subscale “stereotype endorsement”. This pattern suggests that enhanced illness awareness may increase sensitivity to apparent social devaluation while simultaneously enabling cognitive distancing from stereotypes. This evidence reinforces contemporary recovery models, asserting that increased illness awareness without parallel stigma reduction may negatively influence subjective recovery outcomes. Moreover, this illustrates the need for integrating stigma-focused psychological support alongside psychoeducation in early psychosis services.

Social anxiety and avoidance closely paralleled internalized stigma levels, indicating that fear of negative evaluation may represent a behavioral manifestation of self-stigma.In a study conducted by Brain et al. (2014), the authors report that more than half of the participants felt socially excluded. The most common coping strategy they used to bypass stigma and discrimination was avoidance. Almost all concealed diagnoses, avoided close relationships and did not apply for a job. Lysaker et al. (2007) and Park et al. (2013) did not find any correlations between negative symptoms and stigma levels in individuals with chronic schizophrenia, while Rector et al. (2005) report an association between internalised stigma and negative symptoms.In this study, we did not find any correlation between negative symptoms either at baseline or at six months’ follow-up. Taken together, stigma-related social avoidance may represent an adaptive but ultimately maladaptive coping response intended to minimise anticipated discrimination during early illness stages.

Baseline positive symptoms demonstrated a small but significant association with internalized stigma, suggesting that acute symptom expression may raise vulnerability to stigmatizing experiences.Ho et al. (2018) report that more severe positive symptoms at intake predicted higher self-stigma levels at follow-up, and Tarrier et al. (2007) found that the severity of positive symptoms is associated with higher perceived stigma. More severe positive symptoms may lead to agreement with stereotypes that persons with psychosis are unpredictable, unreliable, aggressive or peculiar and to internalisation of social stigma.

Furthermore, Ho et al. (2018) report a significant positive correlation between higher stigma levels and longer DUP and suggest that longer exposure to public stigma may lead to increased internalized stigma. No association between DUP and internalized stigma was observed in the present sample. The absence of association with DUP may reflect the relatively short untreated duration in this cohort, limiting exposure to public stigma during early illness stages. This evidence suggests that subjective interpretations of illness experiences, rather than duration of untreated psychosis alone, may play a more prominent role in early stigma formation.

Finally, the results of the present study suggest that stigma experiences are not associated with gender. Ho et al. (2018) report an association between female gender and higher stigma levels in first episode patients, and Jerkins and Carpenter-Song (2009) report an association between female gender and higher stigma levels in a sample of chronic schizophrenia patients. On the contrary, Livingston & Boyd (2010), in a meta-analysis conducted in a sample of persons with chronic psychiatric disorders, did not find gender differences in stigma levels.These differences across studies may indicate that contextual and ethnocultural factors, rather than gender itself, influence stigma experiences during early psychosis.

Despite these clinically meaningful findings, several methodological limitations should be acknowledged. Participants were recruited from a single specialized early intervention service; therefore, findings may primarily generalize to patients receiving care within structured early intervention services. The cross-sectional design precludes causal inference regarding the relationship between stigma and psychosocial outcomes. The relatively modest sample size from a single early intervention service may limit generalizability. Self-report measures may introduce response bias, even when validated instruments are used. Finally, longitudinal studies are needed to clarify the temporal relationships among insight, stigma, and recovery outcomes.

5. Conclusions

The present findings indicate that internalized stigma represents a key determinant of psychosocial functioning in first-episode psychosis and is closely linked to reduced quality of life and elevated social anxiety. These outcomes support the combination of systematic stigma assessment and targeted anti-stigma interventions as integral components of recovery-oriented early psychosis services.